检索结果-内蒙古大学图书馆

14th International Conference on Frontiers of Information Technology (FIT)

作者： Mahrnood, Zahid Nazeer, Muhammad Arif, Muhammad Shahzad, Imran Khan, Fahad Ali, Mazhar Khan, Uzair Khan, Samee U. COMSATS Inst Informat Technol Dept Elect Engn Abbottabad Pakistan COMSATS Inst Informat Technol Dept Math Wah Cantt Pakistan COMSATS Inst Informat Technol Dept Comp Sci Abbottabad Pakistan North Dakota State Univ Dept Elect & Comp Engn Fargo ND USA

ISBN: (纸本)9781509053001

Recently, large number of object detection and recognition algorithms are playing a key role in several applications, such as security and surveillance. Although, these algorithms perform exceptionally well under normal lighting conditions, however their detection and recognition accuracy abruptly degrades under non-uniform illuminations, such as strong sunlight and bad lighting conditions. In this paper, we apply our own developed Multi-Scale Retinex (MSR) algorithm as a pre-processing module to boost the accuracy of the AdaBoost algorithm, which is considered to be state-of-the-art algorithm for robust object detection and recognition. Simulation results show that the MSR can be reliably and effectively used under non-uniform illuminations to boost the accuracy of the AdaBoost.

关键词： AdaBoost Multi-Scale Retinex object detection and recognition

来源：评论

学校读者我要写书评

暂无评论

An efficient object detection mechanism with LSTM-based object recognition for computer night vision images in edge and cloud environments

引用

APPLIED SOFT COMPUTING 2025年 169卷

作者： Prabu, V. Charles Perumal, Pandiaraja Vel Tech Rangarajan Dr Sagunthala R&D Inst Sci & T Dept Comp Sci & Engn Chennai 600062 Tamil Nadu India

object detection is the critical part for system considering the surveillances, images and so on. Traditionally, the object detection is being helped to identify the objects present in the source to get the related information and used for various applications. Nevertheless, the earlier detection models are focused on images with normal illumination level, when object detection in low illumination images is quite ignored. object detection in images with high illumination is a simple task, and most of the research is also done in that field. However, identifying and locating the objects in night vision or images with low illumination is complex. Due to the advancement, the deep learning method is emerging in imaging technology and application. Yet, it hinders to achieve the same performance while detecting the night vision images due to bad light and low illumination condition. Therefore, a new night vision object detection system is executed in this work. To initiate this work, the mandatory images for conducting the research are gathered from the benchmark dataset. After that, the gathered images are enhanced by the multi-scale retinex network. These enhanced images are given to the object detection phase. Here, the object detection is carried out by the invented 3D Swin Transformer-based Yolov7 structure and the recognition is done using the Long Short-Term Memory (LSTM) approach. Lastly, the outcome achieved from the 3D Swin Transformer-based Deep Learning for object detection and recognition Network (3DST-DL-ODRNet) is correlated with the classical object detection and the recognition system to verify the efficacy of the proposed model. Finally, the proposed system is examined with various measures for detection as well as recognition. Also to ensure the efficiency, the proposed framework is to be compared with classical approaches. On the analysis, the high value of IoU and mAP is achieved as 0.99 and 94.03 for dataset 1, whereas 0.96 and 94.77 for dataset 2. S

关键词： object detection and recognition Computer night vision Edge and cloud environment Multi-scale retinex 3D swin transformer-based deep learning for object detection and recognition network Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Traffic Lights detection and recognition with New Benchmark Datasets Using Deep Learning and TensorFlow object detection API

引用

TRAITEMENT DU SIGNAL 2022年第5期39卷 1673-1683页

作者： Kilic, Irfan Aydin, Galip Firat Univ Dept Comp Engn TR-23119 Elazig Turkey

Today, traffic lights are widely used in places with high vehicle traffic. Especially in autonomous vehicles, fast and high accuracy detection and recognition of traffic lights are critical. Machine learning methods are generally used to do this. Deep learning models give more successful results than machine learning methods in detecting the exact location of traffic lights in different climatic conditions. In this study, Faster R-CNN Inception v2 deep learning model was trained and tested on two different datasets that we prepared and published publicly under variable traffic and climatic conditions in Turkey. Successful results were obtained with fewer data by using the Transfer Learning method with the help of TensorFlow object detection API in the training of the model. It has been shown that the datasets we have prepared can be developed considering the conditions in other countries and successful results will be obtained.

关键词： traffic lights deep learning benchmark datasets TensorFlow object detection API object detection and recognition Faster RCNN

来源：评论

学校读者我要写书评

暂无评论

PF-YOLO:An Improved YOLOv8 for Small object detection in Fisheye Images

引用

Journal of Beijing Institute of Technology 2025年第1期34卷 57-70页

作者： Cheng Zhang Cheng Xu Hongzhe Liu Beijing Key Laboratory of Information Service Engineering Beijing Union UniversityBeijing 100101China Beijing Qiangqiang Yuanqi Technology Co. LtdBeijing 101121China

Top-view fisheye cameras are widely used in personnel surveillance for their broad field of view,but their unique imaging characteristics pose challenges like distortion,complex scenes,scale variations,and small objects near image *** tackle these,we proposed peripheral focus you only look once(PF-YOLO),an enhanced YOLOv8n-based ***,we introduced a cutting-patch data augmentation strategy to mitigate the problem of insufficient small-object samples in various ***,to enhance the model's focus on small objects near the edges,we designed the peripheral focus loss,which uses dynamic focus coefficients to provide greater gradient gains for these objects,improving their regression ***,we designed the three dimensional(3D)spatial-channel coordinate attention C2f module,enhancing spatial and channel perception,suppressing noise,and improving personnel *** results demonstrate that PF-YOLO achieves strong performance on the challenging events for person detection from overhead fisheye images(CEPDTOF)and in-the-wild events for people detection and tracking from overhead fisheye cameras(WEPDTOF)*** to the original YOLOv8n model,PFYOLO achieves improvements on CEPDTOF with increases of 2.1%,1.7%and 2.9%in mean average precision 50(mAP 50),mAP 50-95,and *** WEPDTOF,PF-YOLO achieves substantial improvements with increases of 31.4%,14.9%,61.1%and 21.0%in 91.2%and 57.2%,respectively.

关键词： fisheye object detection and recognition small object detection deep learning

来源：评论

学校读者我要写书评

暂无评论

Enhanced Multi-Pill detection and recognition Using VFI Augmentation and Auto-Labeling for Limited Single-Pill Data

引用

IEEE ACCESS 2025年 13卷 60859-60878页

作者： Lee, Seung-Hwan Son, Dong-Min Lee, Sung-Hak Kyungpook Natl Univ Sch Elect & Elect Engn Daegu 41566 South Korea

This study presents a method for object detection and recognition to identify the positions and types of pills in images containing multiple pills using a small-scale dataset of single-pill images for training. The proposed approach aims to detect multiple pills at the final stage despite the initial training data, which includes only single pills. The method consists of three primary steps. First, a data augmentation technique is introduced to prevent overfitting and improve learning efficiency. This augmentation uses a video frame interpolation (VFI) technique based on the latent diffusion model (LDM). A capturing system is developed for this purpose, and differences between images are used as additional information in weight maps to train the LDM. Second, an automatic labeling system is proposed to generate label data for the paired dataset efficiently. Accurate labeling requires the position and type of pills as training data, but manually labeling the augmented dataset of 61,440 images would be costly. Therefore, an automatic labeling system using an attention map and a deep U-Net is proposed to generate the label data efficiently. Third, a method is presented to detect the position and type of multiple pills based on a training dataset containing only single-pill images. Reliable detection and recognition of multiple pills usually require datasets containing various pill combinations. However, as the number of classes increases, the possible combinations grow exponentially. To address this, we propose a system that learns from single-pill images to detect multiple pills accurately. This study uses a dataset containing 40 types of pills for experimentation, and the results demonstrate superior precision, recall, individual pill accuracy, and image accuracy compared to other methods.

关键词： object detection and recognition data augmentation video frame interpolation automatic labeling system object detection and recognition data augmentation video frame interpolation automatic labeling system

来源：评论

学校读者我要写书评

暂无评论

Concealed objects detection and recognition in active transmission sub-terahertz imaging scanner for automatic security screening

引用

OPTICAL ENGINEERING 2024年第3期63卷 033102-033102页

作者： Hocine, Medjadba Hamid, Zehaf Mohamed, Lazoul Ecole Mil Polytech Lab Syst Optoelect Algiers Algeria Ecole Mil Polytech Lab Syst Lasers Algiers Algeria

As part of developing a sub-terahertz imaging scanner for automatic security screening applications, we investigate the detection and recognition of concealed objects in sub-terahertz images using various image processing techniques capable of addressing the low quality of such images. First, we propose a method that combines locally adaptive and multilevel thresholding with connected components labeling. This method enables the removal of the nonuniform background, facilitating the detection and localization of multiple concealed objects. Subsequently, we analyze the performance of several local feature-based matching methods for selecting the optimal detector-descriptor combination, allowing for efficient detection and recognition of concealed objects. Furthermore, we assess the applicability and performance of the supervised machine learning method based on the support vector machine classifier combined with the bag of visual words model. Finally, we explore the effectiveness of transfer learning by fine-tuning two pretrained convolutional neural networks, namely AlexNet and YOLOv3 models. The two fine-tuned models achieve recognition accuracies of 87% and 92% for AlexNet and YOLOv3, respectively. Benchmarking the study results demonstrates that deep learning methods provide an efficient solution for automatic security screening systems. (c) 2024 Society of Photo-Optical Instrumentation Engineers (SPIE)

关键词： sub-terahertz linear scanner sub-terahertz images object detection and recognition security screening

来源：评论

学校读者我要写书评

暂无评论

Generative Data Augmentation for Automatic Meter Reading Using CNNs

引用

IEEE ACCESS 2022年 10卷 28471-28486页

作者： Sripanuskul, Nuntida Buayai, Prawit Mao, Xiaoyang Univ Yamanashi Fac Engn Kofu Yamanashi 4000015 Japan Hangzhou Dianzi Univ Sch Comp Sci & Technol Hangzhou 310013 Peoples R China

While smart meters are still not widely installed in many countries, automatic reading of traditional-type meters is useful from the perspective of both cost and safety. Although convolutional neural network (CNN) showed a high potential for automatic meter reading under unconstrained environment, it is facing various challenges. One is the difficulty of collecting a sufficient amount of training dataset since some digits of a meter may take a long time to update. Another challenging issue is how to recognize the transitional state between two consecutive numbers. To solve these problems, we propose a new data augmentation technique that can automatically generate annotated images of numbers, including the transitional states. By taking advantage of the state-of-the-art generative neural network model, the generated numbers resemble the local appearance of those in the original meter images. Evaluation experiments confirm that our proposed generative data augmentation techniques improve the robustness of the recognition model and achieve outstanding results when compared to the previous work.

关键词： Meters Training data Neural networks Licenses Image recognition Convolutional neural networks Automatic meter reading Automatic meter reading image data augmentation generative neural network object detection and recognition

来源：评论

学校读者我要写书评

暂无评论

A survey of modern deep learning based object detection models

引用

DIGITAL SIGNAL PROCESSING 2022年第0期126卷

作者： Zaidi, Syed Sahil Abbas Ansari, Mohammad Samar Aslam, Asra Kanwal, Nadia Asghar, Mamoona Lee, Brian Technol Univ Shannon Midlands Midwest Software Res Inst Athlone N37 HD68 Westmeath Ireland Univ Chester Fac Sci & Engn Chester Cheshire England Natl Univ Ireland Insight Ctr Data Analyt Galway Ireland Kee Univ Sch Comp & Math Keele Staffs England Lahore Coll Women Univ Lahore Pakistan Natl Univ Ireland Sch Comp Sci Galway Ireland

object detection is the task of classification and localization of objects in an image or video. It has gained prominence in recent years due to its widespread applications. This article surveys recent developments in deep learning based object detectors. Concise overview of benchmark datasets and evaluation metrics used in detection is also provided along with some of the prominent backbone architectures used in recognition tasks. It also covers contemporary lightweight classification models used on edge devices. Lastly, we compare the performances of these architectures on multiple metrics. (C) 2022 Elsevier Inc. All rights reserved.

关键词： object detection and recognition Convolutional neural networks (CNN) Lightweight networks Deep learning

来源：评论

学校读者我要写书评

暂无评论

Evaluating FPGA Acceleration on Binarized Neural Networks and Quantized Neural Networks 25

Evaluating FPGA Acceleration on Binarized Neural Networks an...

引用

International Symposium on Measurement and Control in Robotics (ISMCR)

作者： Surapally, Sarala K. Yang, Xiaokun Harman, Thomas L. Shih, Liwen Univ Houston Clear Lake Engn Dept Houston TX 77058 USA

ISBN: (数字)9781665454964

ISBN: (纸本)9781665454964

The gap between complex deep learning models and the limited computing capability of individual robots is one of the driving forces to hardware acceleration of image classification networks. Therefore, this paper presents two case studies using the field-programmable gate array (FPGA) acceleration on deep neural networks. The hardware platform is mainly composed of an AMD-Xilinx PYNQ-Z2 board as a computing device and a USB camera as the image/video sensor. The main applications in image/video detection and recognition of traffic signs and objects are performed in this work. Compared with the software computation, the results of the FPGA implementation show a significant speedup to the single road sign recognition using the Binarized Neural Networks and >= 50x speedup to the object detection using Quantized Neural Networks. This paper focuses on the preliminary research into hardware acceleration on neural networks. It demonstrates a great potential to improve the speed for image processing with FPGA, particularly to the time-constrained scenarios. The final goal of this project is to provide parameterizable designs for complicated neural networks with FPGA.

关键词： Binarized Neural Networks FPGA object detection and recognition Quantized Neural Networks

来源：评论

学校读者我要写书评

暂无评论

Characterization Parameter of Material Information in Passive Millimeter-wave Sensing 9

Characterization Parameter of Material Information in Passiv...

引用

2022 International Wireless Symposium

作者： Hu, Yan Su, Jinlong Hu, Fei Wu, Hongfei Huazhong Univ Sci & Technol Sch Elect Informat & Commun Wuhan Hubei Peoples R China

ISBN: (数字)9781665481977

ISBN: (纸本)9781665481977

Passive millimeter-wave (MMW) sensing has been used for several close-range applications such as security checks, military detection, terrain modeling, and so on. By measuring the electromagnetic energy radiated by different materials, it is possible to distinguish them. However, the brightness temperature (TB) can not be directly used for material classification, and those TB-derived material discriminators are not stable characterization parameters. In this article, we analyze the multi-polarization brightness temperature model and propose the equivalent permittivity (EP) to characterize the material information. This parameter, to some extent, is not affected by incident angle and can be derived by multi-polarization measurement. The significant advantage of EP is that it reduces the dimension of parameters in the model. Therefore, it is conducive to extracting object information based on the MMW radiation model. As an application example, we used EP in this article for object classification. Based on EP, we successfully classified four objects with an accuracy of about 99%. Future applications of our method include liquid analysis and scene monitoring.

关键词： millimeter wave radiation object information inversion object detection and recognition 3D information acquisition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：